Home › Directories › Free green AI courses
Directory
Free green AI courses
Free courses and training on green AI, sustainable computing and AI for climate, checked on each provider's own site.
About this directory
25 free courses and learning resources on green and sustainable AI, from beginner introductions to advanced material for developers and researchers. Each one was checked on the provider's own site; "free to audit" means the learning is free but a certificate costs money.
Green Software for Practitioners (LFC131)
Linux Foundation Training & Green Software Foundation
A vendor-neutral introduction to building, running and maintaining software that uses less energy and emits less carbon. Aimed at any software practitioner and their managers.
self-paced · about 2 hours · English; Japanese version (LFC131-JP)
Go to the course →The Principles of Sustainable Software Engineering
Microsoft Learn
A short module on the six core principles of green software: carbon efficiency, energy efficiency, carbon awareness, hardware efficiency, measurement and climate commitments.
self-paced · about 25 minutes (10 units) · English, with many translated versions
Go to the course →Describe the sustainability practices in datacenters
Microsoft Learn
Explains why data centres, the infrastructure behind cloud and AI services, matter for emissions, and how operators use renewable energy and efficiency measures to cut their footprint. Written for students.
self-paced · about 42 minutes (12 units) · English
Go to the course →Introduction to project FarmVibes.AI
Microsoft Learn (Microsoft Research)
Introduces an open-source Microsoft Research toolkit that applies AI to farm data such as soil and crop health. Shows one way AI can support sustainable agriculture and conservation.
self-paced · about 16 minutes · English
Go to the course →AI and Environmental Sustainability
Lund University (Coursera)
Looks at AI's environmental footprint across its whole lifecycle, including energy, materials and e-waste, along with questions of environmental justice and how AI's impact can be governed.
self-paced · about 8 hours (4 modules) · English
Go to the course →Sustainability in the Digital Age: AI and Sustainability – Balancing Innovation and Responsibility
openHPI (Hasso Plattner Institute, with Stanford University and SAP)
Covers how AI can help and harm sustainability goals, including energy-efficient machine learning, greener data centres and ethical and legal issues, with industry case studies.
self-paced · about 5–6 hours over two weeks · English
Go to the course →Sustainability in the Digital Age: Environmental Impacts of AI Systems
openHPI (Hasso Plattner Institute, with Stanford University and SAP)
A non-technical course on measuring and reporting the carbon footprint of AI models. It covers energy and carbon accounting, why estimates are hard, and ways to reduce impact.
self-paced · about 6 hours · English
Go to the course →Sustainability in the Digital Age: Efficient AI Techniques in the LLM Era
openHPI (Hasso Plattner Institute)
Teaches techniques that make deep learning use less compute and energy, such as model compression, low-bit quantisation of large language models and collaborative inference. Requires Python and basic deep learning knowledge.
self-paced · about 6–8 hours over two weeks · English
Go to the course →Sustainability in the Digital Age: Energy-Efficient Software Development
openHPI (Hasso Plattner Institute, with SAP)
Covers sustainable programming, calculating software's carbon footprint and greener operations, including how to measure the performance and energy use of large language models.
self-paced · about 3–4 hours · English
Go to the course →Sustainable Software Engineering
openHPI (Hasso Plattner Institute)
Introduces how IT both saves and causes CO2 emissions, and how choices of architecture, algorithms, data structures, storage and caching affect software's energy use.
self-paced · 3–6 hours per week over two weeks · English; German version available
Go to the course →Energy Efficient Programming
openHPI (Hasso Plattner Institute)
A low-level course on how CPU architecture affects energy use, with C and assembly techniques, intrinsics and hands-on energy measurement using Intel VTune. For computing students and developers.
self-paced · English
Go to the course →clean-IT: Towards Sustainable Digital Technologies
openHPI (clean-IT Initiative)
An open forum-style collection of expert talks on cutting the energy demand of computing, from data centres to AI. There are no deadlines or graded tests, and learners can add their own contributions.
video series · English
Go to the course →Green AI Foundations: Designing Energy-Efficient AI Systems
ED4S
A short crash course for technical leaders on building energy efficiency into AI work at five layers: strategy, infrastructure, model design, application architecture, and governance and measurement.
self-paced · about 30 minutes · English
Go to the course →Carbon Aware Computing for GenAI Developers
DeepLearning.AI with Google Cloud
Shows how to measure computing emissions and use live electricity-grid carbon data to train machine learning models in cleaner cloud regions and at cleaner times, using Google Cloud tools.
self-paced · about 2 hours · English
Go to the course →AI Efficiency Courses
Pruna AI
Open slides and notebook exercises on making AI models smaller and faster, covering quantisation, pruning, distillation, compilation and how to evaluate the trade-offs. Some exercises need a GPU.
self-paced · English
Go to the course →TinyML and Efficient Deep Learning Computing (MIT 6.5940, Fall 2024)
MIT HAN Lab
Recorded lectures and slides from MIT's graduate course on efficient AI, covering pruning, quantisation, neural architecture search, distributed training and efficient large language and diffusion models.
video series · English
Go to the course →Sustainable AI (chapter in Machine Learning Systems, Volume II)
Harvard University (Vijay Janapa Reddi, open textbook)
A textbook chapter on the energy and environmental limits of machine learning systems, from the power budgets of data centres to the scale of AI's footprint and ways to reduce it.
reading · English
Go to the course →Climate Change AI Tutorials
Climate Change AI
A library of over 40 hands-on Python notebooks, many with walkthrough videos, applying machine learning to climate problems such as flood mapping, deforestation detection, weather forecasting and energy systems. Levels range from introductory to advanced.
self-paced · English
Go to the course →Climate Change AI Virtual Summer School 2024 (recordings)
Climate Change AI
Recorded lectures and coding tutorials on using AI for climate action across energy, land use, weather, policy and cities, plus sessions on AI's own environmental footprint and regulation.
video series · 7 weeks, about 4 hours per week · English
Go to the course →Environmental impacts of digital technologies (Impact'Num)
Inria and Class'Code (FUN MOOC)
A general-audience course on the energy and resource footprint of digital devices and services, how to read the indicators, and what citizens, designers and teachers can do. It does not focus on AI specifically.
self-paced · about 5 hours · English and French
Go to the course →Sustainability and AI
University of Helsinki, Ruralia Institute (DigiCampus)
Four short lectures on what sustainable AI means, what AI needs to run and what drives its environmental footprint. Aimed particularly at small-business owners and entrepreneurs.
self-paced · four video lectures plus an optional assignment · English
Go to the course →Introduction to Machine Learning for Earth Observation
EO College (with TUM, HZDR and University of Twente)
Six expert lessons on applying machine learning to satellite data, including ethics, uncertainty, self-supervised learning and machine learning for the UN Sustainable Development Goals.
self-paced · about 7 hours · English
Go to the course →Experience AI: AI and the environment
Raspberry Pi Foundation and Google DeepMind
Free classroom materials for teachers, including a flood-forecasting unit for ages 8–12 and an ecosystems lesson for ages 11–14. Pupils explore the environmental footprint of AI and how AI supports conservation.
self-paced · English, with localised versions in many languages
Go to the course →CIRA Short Course on Machine Learning for Weather and Climate
Cooperative Institute for Research in the Atmosphere (CIRA), Colorado State University
Six recorded lectures with notebooks taking atmospheric scientists from regression and decision trees through to neural networks, CNNs and model interpretation, using weather and climate examples.
video series · six 90-minute lectures · English
Go to the course →Computational Tools for Climate Science
Climatematch Academy
An open course book of Python tutorials on climate data, from reanalysis and remote sensing to future climate projections, including a day on how machine learning can help climate modelling.
self-paced · English
Go to the course →Frequently asked questions
Are these courses really free?
Yes. Every course here can be followed without paying. Some platforms charge for a certificate or graded assessments; those are marked "free to audit".
How do courses get listed?
We list free courses that cover green or sustainable AI, energy-efficient computing or AI for climate, after checking them on the provider's own site. Anyone can suggest a course; listing courses is free.